DuPont Analysis: Breaking Down Return on Equity

DuPont analysis breaks return on equity into the three (or five) levers that actually drive it: profitability, asset efficiency, and leverage. Two companies can post an identical 25% ROE for completely different reasons - one earns it through fat margins, the other through debt - and DuPont is the formula that tells you which is which.

Return on equity (ROE) is the single most-quoted profitability metric in corporate finance, because it answers the question every shareholder actually cares about: for every dollar of equity I've put in, how much profit did the business generate this year? The formula is simple - Net Income divided by Shareholders' Equity - which is also its weakness. A high ROE looks great on a slide, but the raw ratio can't tell you whether it came from strong operations or from leverage doing the heavy lifting.

DuPont analysis, developed by the finance department at E.I. du Pont de Nemours and Company in the 1920s, fixes that by decomposing ROE into its component drivers. The classic version splits ROE into three multiplicative pieces: net profit margin, asset turnover, and the equity multiplier (a measure of leverage). Multiply them together and you get back to ROE exactly - but now you can see which lever is actually moving the number.

The 3-Step DuPont Formula: profitability × efficiency × leverage = ROE

The Three-Step DuPont Formula

The classic (3-step) DuPont identity is:

ROE = Net Profit Margin × Asset Turnover × Equity Multiplier

Each term isolates a different question about how the business generates its return:

Multiply the three together and the Revenue and Total Assets terms cancel out algebraically, leaving Net Income / Equity - which is just the ROE formula. That's the whole trick: DuPont doesn't change the answer, it just shows you the path to it.

// Net Profit Margin
= Net_Income / Revenue

// Asset Turnover
= Revenue / Total_Assets

// Equity Multiplier
= Total_Assets / Total_Equity

// ROE (3-step)
= Net_Profit_Margin * Asset_Turnover * Equity_Multiplier

A business can push ROE higher through any one of these three levers: fatten margins, sell more per dollar of assets, or lean on more debt. Only the first two reflect operating improvement - the third is financial engineering, and it's the one that turns a mediocre business into a scary one when it goes wrong. That distinction is the entire reason DuPont exists as a separate exhibit rather than just reporting ROE on its own - see our guide on how the three financial statements link together for how net income, total assets, and equity flow between the income statement and balance sheet in the first place.

The Five-Step (Extended) DuPont Formula

The 3-step version treats "net profit margin" as one number, but margin itself is a function of three separate things: how much interest a company pays, how much tax it pays, and how profitable its core operations are before either. The extended (5-step) DuPont formula splits net margin apart to isolate each:

ROE = Tax Burden × Interest Burden × EBIT Margin × Asset Turnover × Equity Multiplier

Tax Burden × Interest Burden × EBIT Margin multiplies back out to Net Income / Revenue - the same net margin figure from the 3-step formula - so the 5-step model is strictly more granular, not a different answer. Its value is diagnostic: it tells you whether a margin decline is coming from operations (EBIT margin falling), from a heavier debt load (interest burden rising), or from a change in tax rate - three very different problems that a single "net margin fell" line would blur together.

// Tax Burden
= Net_Income / EBT

// Interest Burden
= EBT / EBIT

// EBIT Margin
= EBIT / Revenue

// ROE (5-step)
= Tax_Burden * Interest_Burden * EBIT_Margin * Asset_Turnover * Equity_Multiplier

Worked Example: Decomposing a 25% ROE

Take a company with the following current-year financials:

At first glance, ROE = $40M / $160M = 25.0% - a strong return by almost any benchmark. DuPont tells you why.

3-step decomposition:

5-step decomposition (same company, same year):

Both paths reconcile exactly to the direct calculation (Net Income / Equity = $40M / $160M = 25.0%), which is the built-in check every DuPont model should carry: if your decomposed ROE doesn't tie back to the one-line ratio, a formula is wrong somewhere upstream.

The 5-step view adds real information here: EBIT margin of 14.0% is healthy, and the interest burden of 85.7% shows the company is keeping most of its operating profit after debt service - leverage is being used moderately, not aggressively. If interest burden had instead been 60%, that would flag a company whose debt load is eating a much larger share of operating profit, well before it shows up as a covenant problem.

DuPont measures return to equity holders specifically. If you also want to sanity-check the return on a specific investment or project independent of how it's financed, that's a job for a plain return-on-investment calculation rather than ROE:

Peer Benchmarking: The Same ROE, Different Drivers

A 25% ROE means nothing in isolation - the real value of DuPont shows up when you line a company up against its peers. Here's the same company benchmarked against four comparable peers in the current year:

The subject company beats the peer average ROE by 4.3 percentage points (25.0% vs. 20.7%). DuPont shows exactly where that edge comes from: net margin is 0.9 points above the peer average and the equity multiplier is 0.27x higher than peers - the company is both somewhat more profitable per sales dollar and running with more leverage than its peer set, while asset turnover is actually a touch below average (1.25x vs. 1.33x). An analyst reading only the headline ROE would call this company "better than peers." An analyst reading the DuPont breakdown would say it's better and more levered than peers - a materially different risk conclusion for the same number.

Reading a Trend: When Flat ROE Hides a Problem

DuPont is at its most useful when you run it across several years, because a stable ROE can mask a deteriorating business. Consider this three-year trend for a single company:

Headline ROE looks essentially flat - 23.4% in Year 1, 23.4% again in Year 3. A reader who only tracks the ROE line would conclude nothing changed. But every operating driver moved in the wrong direction: net margin fell from 9.0% to 7.0% and asset turnover slowed from 1.30x to 1.20x. The only reason ROE held steady is that the equity multiplier climbed from 2.00x to 2.79x - the company took on progressively more leverage to offset a weakening operating picture. That's precisely the pattern credit analysts and short-sellers screen for: rising leverage propping up a return metric while the underlying business quietly deteriorates.

Common Mistakes

1. Treating ROE as a single number without decomposing it. A 25% ROE from margin and efficiency is a fundamentally different (and safer) business than a 25% ROE propped up by 5x leverage. Never compare two companies' ROE without checking the DuPont drivers first.
2. Mixing book equity with market equity. DuPont uses book values from the balance sheet (Total Equity, Total Assets) consistently. Substituting market capitalization for book equity in one part of the formula while using book figures elsewhere breaks the algebraic identity and produces a number that doesn't reconcile.
3. Using average balances inconsistently. Some analysts use average total assets and average equity (beginning + ending, divided by two) instead of ending balances, which is more accurate for turnover ratios but must be applied consistently across every driver - mixing average and ending balances is a common source of a decomposition that doesn't tie back to the direct ROE calculation.
4. Ignoring negative equity. If a company has negative shareholders' equity (common after a large buyback or years of losses), the equity multiplier and ROE become mathematically meaningless or wildly misleading. Flag this case rather than reporting a nonsensical ROE.
5. Comparing ROE across industries without adjusting for structural leverage. Banks and REITs run naturally high equity multipliers because of the nature of their business; comparing their ROE directly to a software company's ROE without accounting for that structural difference in the equity multiplier is comparing apples to oranges.
6. Not reconciling the decomposition back to the direct ratio. Every DuPont model should include a check row: decomposed ROE minus (Net Income / Equity) should equal zero. If it doesn't, there's a formula or a data error somewhere in the chain.